Interpretable Machine Learning of Nanoparticle Stability through Topological Layer Embeddings
Felipe Hawthorne, Leandro Seixas, James M. Almeida, Cristiano F. Woellner, Raphael M. Tromer · 2026
Read Interpretable Machine Learning of Nanoparticle Stability through Topological Layer Embeddings by Felipe Hawthorne, Leandro Seixas, James M. Almeida, Cristiano F. Woellner, Raphael M. Tromer (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-17528.txt.
Description
Felipe Hawthorne led a five-author team behind a 2026 arXiv paper on nanoparticle stability. Stability turns on a huge space of atomic arrangements across varied surface and interior settings. Hunting for the lowest energy shapes by first principles alone is costly, so the paper offers a framework built to work from only a few hundred reference calculations.
At its core sits a layer-resolved descriptor that splits each particle into surface, middle, and core zones. It keeps a compact feature size while marking where each atom sits. Weighting schemes let the model stress one zone over another, and gradient-boosted trees paired with ranking then pick out the most stable shapes. Tests show near-saturated ranking correlation, strong top-rank recall, and regret that fades fast. Layer weighting and SHAP analysis show how surface segregation, coordination patterns, and local disorder shape stability zone by zone.
Catalog record
| CREATOR | Felipe Hawthorne, Leandro Seixas, James M. Almeida, Cristiano F. Woellner, Raphael M. Tromer |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 67 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-17528